Search Results

A variety of methodologies to use embedded multicore controllers efficiently has been discussed in the last years. Several assumptions are usually made in the automotive domain, such as static assignment of tasks to the cores. This paper shows an approach for efficient task allocation depending on different system modes. An engine management system (EMS) is used as application example, and the performance improvement compared to static allocation is assessed. The paper is structured as follows: First the control algorithms for the EMS will be classified according to operating modes. The classified algorithms will be allocated to the cores, depending on the operating mode. We identify mode transition points, allowing a reliable switch without neglecting timing requirements. As a next step, it will be shown that a load distribution by mode-dependent task allocation would be better balanced than a static task allocation.

Infineon's latest high-end automotive microcontrollers like TC1796 are complex Systems On Chip (SoC) with two processor cores and up to two internal multi-master buses. The complex interaction between cores, peripherals and environment provides a big challenge for debugging. For mission critical control like engine management the debugging approach must not be intrusive. The provided solution are dedicated Emulation Devices which are able to deal with several 10 Gbit/s of raw internal trace data with nearly no cost adder for mass production and system design. Calibration, which is used later in the development cycle, has different requirements, but is covered by the Emulation Devices as well. The architecture of TC1796ED comprises the unchanged TC1796 silicon layout, extended by a full In-Circuit Emulator (ICE) and calibration overlay memory on the same die. In most cases, the only debug/calibration tool hardware needed is a USB cable.

Integration scenarios for ECU software become more complicated, as more constraints with regards to timing, safety and security need to be considered. Multi-core microcontrollers offer even more hardware potential for integration scenarios. To tackle the complexity, more and more model based approaches are used. Understanding the interaction between the different software components, not only from a functional but also from a timing view, is a key success factor for high integration scenarios. In particular for multi-core systems, an amazing amount of timing data can be generated. Usually a multi-core system handles more software functionality than a single-core system. Furthermore, there may be timing interference on the multicore systems, due to the shared usage of buses, memory banks or other hardware resources.

Multi-core systems are adopted quickly in the automotive domain, Proof of concepts have been implemented for power train, body and chassis, involving hard real-time constraints. However, depending on the degree of integration, it can be costly, especially in those cases where existing single-core software has to be migrated over. Furthermore, there seems to be a high level of uncertainty, whether a found solution, with regards to partitioning, mapping and orchestration of software is close to an optimum solution. Some integrated solutions demonstrate considerably less performance, for instance due to communication overhead compared to execution on single-core systems. This paper discusses a methodology, as to how to effectively and efficiently investigate the software architecture design space for multi-core software development.

The infrastructure in modern cars is a heterogeneous and historically grown network of different field buses coupling different electronic control units (ECUs) from different sources. In the past years, the amount of ECUs in the network has rapidly grown due to the mushrooming of new functions which historically were mostly implemented on a one-ECU-per-function basis resulting in up to a hundred ECUs in fully equipped luxury cars. Additionally, new functions like parking assist systems or advanced chassis control functions are getting increasingly complex and require more computing power. These two facts add up to a complex challenge in development. The current trend to host several functions in single ECUs as integration platforms is one attempt to address this challenge. This trend is supported by the increased computing power of current and upcoming multi-core microcontrollers.

Automotive powertrain and safety systems under design today are highly complex, incorporating more than one CPU core, running with more than 100 MHz and consisting of several 10 million transistors. Software complexity increases similarly making new methodologies and tools mandatory to manage the overall system. The use of accurate virtual prototypes improves the quality of systems with respect to system architecture design and software development. This approach is demonstrated with the example of the PCP/GPTA subsystem for Infineon's AUDO-NG powertrain controllers.

In recent years, we see more and more ECUs integrating a huge number of application software components. This process mostly results from the increasing amount of so called in-house software in various fields like electric-drive, chassis and driver assistance systems. The software development for these systems is partially moved from the supplier to the car manufacturers. Another important trend is the introduction of new network architectures intending to meet the growing communication requirements. For such ECUs the software integration scenarios become more complicated, as more quality of service requirements with regards to timing, safety and security need to be considered [2]. Multi-core microcontrollers offer even more potential variants for integration scenarios. Understanding the interaction between the different software components, not only from a functional, but also from a timing view, is a key success factor for modern electronic systems [6,7,8,9].

End of Line tests are brief set of tests intended to evaluate ECU's in order to ensure correct functioning of its intended functionality. As these tests are executed on the production line, available time to perform these tests is limited. On one hand, faster production demands require these tests and its framework to be designed in a time optimized manner. On the other hand, increase in ECU functionality translates to an increase in test's functional coverage, requiring more time. Therefore the time taken to execute the tests reaches a critical point in overall ECU production. Availability of multicore microcontrollers with increase in clock speed can increase the performance of end of line tests, but design challenges e.g. synchronization do not guarantee a linear performance increase. Therefore, design of test execution framework is absolutely critical to increase performance of test execution.

Software complexity of vehicles is constantly growing especially with additional autonomous driving features being introduced. This increases the risk for bugs in the system, when the car is delivered. According to a car manufacturer, more than 90% of availability problems corresponding to Electronic Control Unit (ECU) functionality are either caused by software bugs or they can be resolved by applying software updates to overcome hardware issues. The main concern are sporadic errors which are not caught during the development phase since their trigger condition is too unlikely to occur or is not covered by the tests. For such systems, there is a need of safe and secure infield diagnosis. In this paper we present a tool software architecture with remote access, which facilitates standard read/write access, an efficient channel interface for communication and file I/O, and continuous trace.